XForms teleconference 13th November 2000


Josef Dietl, Mozquito Technologies (chair)
Dave Raggett, W3C/Phone.com (scribe)
Steven Pemberton, CWI
TV Raman, IBM
Linda Welsh, Intel
Roland Merrick, IBM
Panagiotis Reveliotis, Philips
Roli Wendorf, Philips
Tom Schnetlage, University of Berkeley
Sebastian Schnitzenbaumer, Mozquito Technologies

Task Force Reports

Model: Linda says she and Roland are working on it. The aim is for a new draft at the end of this week.

Linda is going to email Peter for help with the format attribute. We agree to use a reference to the WML spec for this week and to replace this with some new text in place in the future.

UI: Raman is working through the face to face resolutions, and will send in a new draft in just under two weeks time.

Send: no one present.

Action: Steven to send Gavin, Rob and Mike Mansell a reminder to progress work on the XForms protocols.

Architecture: Peter has sent a document but couldn't make this meeting. We defer this to next week.

Minutes of the Face to Face

We formally resolve to accept the minutes of the Munich face to face.

Money Type

There has been several emails on this recently. The namespace issue would be solved if we used a concatenated strings such as "9.99 USD".

Raman is nervous about this approach.

Roland notes that Steven forwarded an email indicating that the Schema Group has shown some interest in defining a money type.

Josef proposes to ask the XML Schema group about the the schema issues (see his recent email).

Action: Josef to write to the XML Schema Group for their comments on namespaces and complex elements as per his today's email to the XForms list

Action: Steven to find out more about Schema's interest in a money type.

In the meantime, we will work on two strawmen for the current draft, one using the structured string and the other using a complex type.

Raman says we ought to retain the currency type. Steven and Roland suggest this may prejudge the solution to the money type.

Dave suggests there is value in being able to ask users what currency they want to use.

We agree to retain a currency type in the next draft.

Dynamic Constraint Language

Micah isn't here today, and Dave summarised his understanding of Micah's email. It covers the use of XPath functions and also the use of nodesets in the expression language. Dave notes that the latter needs the semantics to be clarified. Raman agrees. We are in favor of studying this further.

Steven talks about the possible use of nodesets for operations on rows or columns of data.

Dave is concerned that whatever semantics we adopt, it should be easy for most people to understand.

We discuss second order functions for a while. Dave proposes we stick to a very small set of such functions, e.g. sum, average, min, and max.

Raman proposes we allow people to create a virtual nodeset which you can then sum etc.

Action: Raman and Dave to study how other languages deal deal with second order functions and to report back on email.

Dave is concerned about the implementation overheads in fully supporting XPath and all its libraries on resource constrained devices such as mobile phones.

Action: Dave to report back on whether it's important or not to subset XPath for small devices, and to ask Peter Stark for his feedback on the same question.